Fault Recovery Method for DG-Containing Distribution Networks Taking into Account Load Forecasting

It is important to deal with distribution network faults in a timely and effective manner. In this paper, a method for islanding and reconfiguration of DG-containing distribution networks combined with load forecasting is proposed. Firstly, power supply is restored to the loads in the non-fault area. Secondly, in order to determine the islanding scheme, a quantum particle swarm optimized long and short-term memory neural network is used to predict the power of loads and distributed power sources during the fault time in combination with meteorological conditions. Finally, the objective function of the active distribution network is constructed to find the optimal scheme for fault recovery using a genetic-based proposed Newton hybrid algorithm. The results of the PG&E 69-node system arithmetic simulation show that the method used in this paper can achieve optimal distribution network fault recovery.